Discussion

Transhepatic venous access may be very useful when conventional access sites are untenable, such as in cases with occluded iliofemoral veins or interrupted inferior caval vein. ^{Reference Ebeid1–Reference Shim, Lloyd, Cho, Moorehead and Beekman5}

Transbaffle puncture is commonly needed to treat arrhythmias in patients with Fontan palliation. Transhepatic access provides a trajectory perpendicular to the baffle, allowing a more stable position of the transseptal needle when force is needed to overcome the thickness and stiffness of the baffle. ^{Reference Singh, Neuzil and Skoka4,Reference Piorkowski, Eitel and Rolf6}

We used fluoroscopic guidance to obtain transhepatic access. Alternatively, ultrasound guidance could be used. ^{Reference Ebeid1,Reference Shim, Lloyd, Cho, Moorehead and Beekman5} Placing the wedge catheter in the hepatic vein via a femoral vein served four purposes. First, we used it as a target during transhepatic needle advancement. Second, it allowed us to judge the distance of the sheath from the accessed vein during closure. Third, an angiogram through it confirmed successful closure of the tract. Fourth and probably most important, it could be used for "bailout" if hemostasis could not be achieved: the wedge catheter could be quickly inflated at the junction of the accessed hepatic vein and the parenchymal tract until hemostasis was achieved (alternately, the wedge catheter could be exchanged over a wire for an angioplasty balloon to be used for balloon occlusion to help achieving hemostasis).

Although the need for closure of a small transhepatic tract is debatable, it is widely accepted to close the tract if a large introducer is placed, especially in patients with increased central venous pressure or bleeding diathesis. ^{Reference Ebeid1}

In the field of cardiac catheterisation, the transhepatic tract is usually closed using coils or occlusion devices. ^{Reference Ebeid1,Reference Singh, Neuzil and Skoka4} Gelatin sponge was reported in only one case following atrial septal defect closure using an 8-Fr sheath. ^{Reference Haddad, Maleux, Bonnet and Malekzadeh-Milani3} On the other hand, gelatin sponge “torpedoes” have been commonly used to close transhepatic portal vein tracts for sheaths up to 6-Fr. ^{Reference Saad and Madoff7–Reference Medsinge, Zajko, Orons, Amesur and Santos9} To the best of our knowledge, our case represents the largest transhepatic venous tract that was closed successfully with gelatin sponge “torpedoes.” The Agilis steerable sheath has an inner and outer diameter of 8.5 and 11-Fr, respectively. ^{Reference Piorkowski, Eitel and Rolf6} Hence, we used a 9-Fr sheath in its place to deliver the gelatin “torpedoes,” anticipating that this sheath would have a similar outer diameter of approximately 11-Fr.

The gelatin sponge is able to absorb blood and provide a scaffold for clot forming. Therefore, it has been used primarily to achieve local hemostasis at the oozing surgical sites. Other uses are embolising arterial vessels and occlude transparenchymal tracts after direct organ access, such as liver, spleen, and lung accesses. ^{Reference Medsinge, Zajko, Orons, Amesur and Santos9,Reference Özer and Köstü10} Gelatin sponge has many advantages over coils or occlusion devices for closure of transhepatic tracts. First, relatively small caliber “torpedoes,” such as the ones we packed into the cut 5-Fr sheath, can be delivered within larger caliber tracts because the gelatin sponge expands when it gets in contact with blood. In addition, gelatin sponge gets absorbed within a few weeks and does not create artefacts with certain imaging modalities such as CT or MRI, it is relatively inexpensive, easy to use, and can be traversed with a needle for future transhepatic access. Importantly, if a gelatin “torpedo” is delivered too close to the hepatic capsule and protrudes into the abdominal wall and past the skin, it can be simply trimmed with scissors; in contrast, coils or nitinol-based devices are problematic if they are positioned across the abdominal wall, and cannot be left to protrude through the skin and would instead require removal, potentially resulting in intra-abdominal bleeding. Gelatin sponge material can be used as a slurry or made into “torpedoes.” To employ a slurry, the gelatin sponge is cut into small pieces and mixed with contrast and then injected inside the tract. We favoured the use of dry “torpedoes” which we expected to expand and take the shape of the tract providing more stability. ^{Reference Haddad, Maleux, Bonnet and Malekzadeh-Milani3,Reference Uller, Muller-Wille and Grothues8,Reference Medsinge, Zajko, Orons, Amesur and Santos9,Reference Özer and Köstü10} We considered this important in our patient with multiple risk factors for bleeding, including the use of a large-bore sheath, increased central venous pressure, and being on dual antiplatelet therapy.